KCL DRIVE-Health Seminar - Harmony Meta - Unlocking 5.5M Variables in UK Longitudinal Studies

Recorded on 24 September 2026 at King's College London, Guy's Campus - Gowland Hopkins Lecture Theatre, Hodgkin Building. Bettina Moltrecht, PhD (University College London) and Thomas Wood (Fast Data Science) introduce Harmony, an open-source platform designed to solve major challenges in global mental health research by streamlining data discovery and retrospective questionnaire harmonization. Key Topics Covered: 1. The Challenges in Mental Health Research: Explores how data fragmentation across global cohorts and time-consuming manual data harmonization traditionally slowed down critical research, such as studying pandemic mental health trends or comparing cross-country outcomes (e.g., UK and Brazil). 2.Harmony Meta: A semantic search engine that connects multiple global metadata catalogues and study databases, enabling researchers to search for study variables using large language models instead of basic keyword matching. 3. Retrospective Harmonization: Demonstrates how Harmony automates the comparison of psychological questionnaires (such as the GAD-7) across different studies and formats, cutting a process that once took months down to seconds. 4.Cross-Lingual & Specialized AI Support: Highlights Harmony's ability to handle multilingual surveys (such as English and Portuguese) using custom-trained language models tailored for the mental health domain. Learn more and try the tool at https://harmonydata.ac.uk/search. Hosted by ‪@drdavidtang‬ and Tech Brews London